What problem does it solve? When a requirement changes an existing C# batch backed by SQL Server stored procedures, teams struggle to predict impact across huge input combinations, trace execution paths across both layers, and prove that new behavior matches expectations without running unsafe tests against production. ## Core Features & Use Cases - Deterministic Combination Analysis: Generate Cartesian candidate sets, apply declarative constraints, and build strength-2 pairwise orthogonal tables without sending the full candidate space to the model. - Old/New Result Comparison: Normalize and classify adapter-produced JSON records into baseline_match, planned_difference, unexplained_difference, business_invalid, system_error, uncertain, and not_executed categories. - Source Decision-Table Extraction: Scan .cs and .sql files to extract if/when/case conditions, factors, and literal values with file-and-line evidence, keeping unresolved expressions as explicit uncertainties. - Use Case: A requirement changes a nightly batch's parameter handling. Use this Skill to trace the C#-to-SP execution path, generate a constrained candidate set, compare old and new isolated-database runs, and produce a reduced regression set with human-gated release decisions. ## Quick Start Ask the AI to use the batch-impact-regression skill to analyze your C# batch and SQL Server stored procedure change and produce a combination regression report from your configuration file.
